Modelling Ecosystem Service Outputs Using Habitat Characteristics

By: Roy Haines-Young and Marion Potschin, University of Nottingham

1. Introduction and approach

This BBN is an experimental tool for modelling the level of ecosystem service outputs based on knowledge of selected habitat characteristics. It was developed as part of work in the UK undertaken by Environment Systems and the University of Nottingham for JNCC.

The system takes the Phase I Habitat Classification system used in the UK as the framework for the analysis. Terrestrial and coastal habitats at level 3 in this classification system were characterised in terms of four biophysical parameters:

The article looked at four different scenarios which may be run using the nodes below:

• rooting depth

• below ground species richness

• canopy height

• above ground species richness

Using expert judgement the habitats were then assessed on each of these parameters and the scores used to predict the level of output of six ecosystem services:

Provisioning	Cultivated crops
Cultural	Wild species diversity Physical and experiential
Regulating	Soil carbon storage Vegetation carbon storage Water quality Water quantity

The expert assessment was used to train the network shown above using the HUGIN software.

2. Using the Network

Using the network you can select a habitat at any level in the Phase 1 system, and the BBN will predict the level of service output based on the four biophysical characteristics of the habitat. Alternatively you can select an ecosystem service and by selecting an output level see which habitats are likely to have this level of performance. By selecting the output for a one service you can also track the trade-offs and synergies with the others.

Note: This is an experimental prototype, and the accuracy of the estimates have not been tested. Moreover, the volume of training data is limited and so not all relevant combinations of habitats and biophysical characteristics might be covered. However, the system does show how a web-based tool could be developed and potentially used to capture people.s experience so that it can be shared with others.

Below is a set of HUGIN widgets for interacting with the model (click on the probability bar to instantiate a node or remove evidence):

---

---

Contact information

For further details contact: Roy Haines-Young@Nottingham.ac.uk

---